Independence of Two Conformations of Sarcoplasmic Reticulum Ca2+-ATPase Molecules in Hydrolyzing Acetyl Phosphate

The sarcoplasmic reticulum Ca2+-ATPase molecules have been shown to exist in two conformations (A and B) that result from intermolecular interaction of ATPase molecules (Nakamura, J., and Tajima, G. (1995) J. Biol. Chem. 270, 17350–17354). The A form binds two calcium ions noncooperatively, whereas the B form binds the calcium ions cooperatively. Here, we examined the independence of these two forms in the calcium-activated hydrolysis of acetyl phosphate (AcP) under asynchronous and synchronous conditions of theirE 1-E 2 transitions at 0–5 and 25 °C. Irrespective of their synchronism and temperature, the two forms hydrolyzed AcP due to calcium that was bound to each of the forms, indicating the independence of the two forms in hydrolyzing AcP. Taking into account the monomer-dimer transition of the ATPase molecules on the sarcoplasmic reticulum membrane accompanyingE 1-E 2 transition of the molecules (Dux, L., Taylor, K. A., Ting-Beall, H. P., and Martonosi, A. (1985) J. Biol. Chem. 260, 11730–11743), the two types of molecules seem to independently carry out such monomer-dimer transition of the same type of molecules. Two pairs, each consisting of the same type of molecules, are suggested to be the structural unit of the ATPase molecules.

Ca 2ϩ -ATPase is an integral membrane protein (110 kDa) of the SR 1 and is a calcium pump protein that transports 2 mol of calcium ion across the SR membrane by hydrolytic coupling with 1 mol of ATP (1)(2)(3). Calcium binding to the two calcium transport sites of the ATPase, being required for phosphorylation of ATPase, is a crucial step in driving the calcium pump. The formation of 1 mol of phosphorylated ATPase and its isomerization are coincident with this calcium transport. Relationships between the structure and function of the calciumtransporting ATPase are being actively investigated (4,5). Furthermore, from observations of the calcium sensitivity of intrinsic fluorescence intensity in the ATPase protein molecules and of calcium binding to the molecules, the ATPase molecules have been shown to exist in two conformations that have the same amino acid sequence (6). Here, for the sake of simplicity, we name the two conformations A form and B form, respectively. The two forms independently bind two calcium ions in a different manner (7): the A form noncooperatively binds two calcium ions with a pH-independent affinity for the ions, whereas the B form cooperatively binds the calcium ions with a pH-dependent affinity for the ions. Recently, however, these nonequivalences of the two forms have been shown to be canceled by solubilization of the ATPase molecules in the SR membrane with detergent, suggesting that intermolecular interaction of the ATPase molecules produces such nonequivalences (8). We therefore further examined the independence of these forms in transporting calcium ions by coupling with hydrolysis of a substrate as an energy source. As to the intermolecular interaction of the ATPase molecules in the SR, Dux et al. (9) found "monomer ribbon-dimer ribbon" transition in the array of molecules on the SR membrane, which coincides with E 1 (high affinity state of the molecule for calcium)-E 2 (low affinity state of the molecule for calcium) transition. The transition of the molecules from E 2 into E 1 has been found to be the slowest step of the transport sequence in the absence of a higher concentration of ATP than that required for phosphorylation of the ATPase molecule (10). Moreover, it has been shown that the A form is in pH-dependent equilibrium between E 1 and E 2 , before calcium binding, whereas the B form preexists predominantly in E 2 independent of pH (7). The transition rates of the two forms from E 2 to E 1 were previously found to be the same, independent of temperature (7,21). Therefore, in this study, the enzyme states of the two forms before calcium binding were set at E 1 and E 2 and at E 2 only, respectively, by setting the pH at specified values, to drive their hydrolysis reaction asynchronously and synchronously, respectively. Moreover, the experimental temperature was varied to carry out the experiments under different conditions of stability of the intermolecular interaction of the ATPase molecules. We then examined whether the two forms independently hydrolyze a substrate due to calcium that is bound to each of the forms; the total activity of substrate hydrolysis by the two forms was compared with calcium binding to the forms in the absence of the substrate by varying the calcium concentration. For direct comparison of the activity and the binding, it is necessary to use a substrate that supports calcium transport and that does not have any regulatory effect on the calcium binding and the enzymatic reaction. AcP has been shown to support calcium transport that couples with its hydrolysis (11), like ATP. Unlike ATP, however, it has not been shown to have a regulatory effect on E 1 -E 2 transition in its hydrolysis cycle (12,13). However, its effect on calcium binding is not known. In this study, we first clarified that AcP had no effect on calcium binding and then used it as the substrate. The results show that the two forms hydrolyze AcP due to calcium that is bound to each of the forms, irrespective of temperature and synchronism of their hydrolysis cycles, indicating the independence of the two forms in hydrolyzing AcP. * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Based on the monomer-dimer transition of the ATPase molecules on the SR membrane (9), the structural unit of the molecules is discussed.

Materials
The procedures for isolation of the SR from the skeletal muscle of rabbit were the same as those described in a previous paper (14). Membranous Ca 2ϩ -ATPase was purified from the SR by washing the SR with sodium deoxycholate in the same manner as reported previously (15) in a 1:5 ratio of sodium deoxycholate to the reticulum protein. The maximum calcium binding capacity of the ATPase preparation was 9.0 -10.7 nmol/mg of protein as estimated by Scatchard analysis of the calcium titration curve of the calcium-binding sites in the preparation, and the maximum level of the preparation that was phosphorylated with ATP was 4.2-4.6 nmol/mg of protein obtained in 0.1 mM ATP, 0.1 mM CaCl 2 , and 5 mM MgCl 2 at pH 7.0. The purified ATPase preparation exhibited the same characteristics of calcium binding as those of the ATPase preparation (7) (see Figs. 1 and 2) that was purified by solubilization of the sarcoplasmic reticulum with sodium deoxycholate and by reformation of the membranous ATPase by removal of the detergent (data not shown).

Assays
AcP Hydrolysis-The total AcP hydrolysis activity of the ATPase was carried out in 20 mM PIPES buffer solution (pH 6.23 or 7.40) containing 0.1 or 0.5 mg of ATPase protein/ml, 0.12 M KCl, 5 mM MgCl 2 , 5 mM AcP, and 0.01-1000 M Ca 2ϩ at 5 or 25°C. To obtain reproducible values of the activities at these pH values and temperatures, the reaction times in which the reactions were linear and the protein concentrations of the ATPase were set at 3 (at pH 7.40) or 5 (at pH 6.23) h and 0.5 mg of protein/ml at 5°C and at 30 (at pH 7.40) or 60 (at pH 6.23) min and 0.1 mg of protein/ml at 25°C, respectively, based on the following observations. (i) The AcPase activities at pH 6.23 were about half of those at pH 7.40. (ii) The activities at 5°C were one-thirtieth to one-fortieth of those at 25°C. (iii) No inactivation of the activity was observed when the ATPase was preincubated in the absence of AcP for 2 and 6 h at 25 and 5°C, respectively, which is different from the case (8) of the detergentsolubilized ATPase, which was gradually inactivated during the preincubation. The calcium-independent activity of AcP hydrolysis was determined under identical conditions, except that 5 mM EGTA was added without the addition of calcium. The calcium-dependent activity of AcP hydrolysis (Ca 2ϩ -AcPase activity) was obtained by subtracting the calcium-independent activity from the total activity. The amount of remaining AcP was determined according to the method of Lipmann and Tuttle (16). Before reading the absorbance at 505 nm of the brown complex formed with iron(III), the ATPase protein suspended in the medium for the assay of AcP was removed by centrifugation (1000 ϫ g for 5 min).
Calcium Binding-Calcium binding was performed according to the rapid filtration method (17) at 0°C as described previously (18). A Millipore HA filter (0.45-m pore size) was used. The ATPase preparation (0.2 mg of protein/ml) was preincubated in 20 mM PIPES buffer solution (pH 6.23 or 7.40) containing 0.12 M KCl and 5 mM MgCl 2 FIG. 1. Calcium binding in the absence of AcP at pH 7.40 and 0°C (7). A, calcium titration curves of the two types of ATPase molecules. E, total binding to the A and B forms; å, rapid binding to the A form, which is in E 1 /E 2 dependent on pH and pre-exists in E 1 at this alkaline pH; Ç, slow binding to the B form, which pre-exists in E 2 independent of pH (see "Results and Discussion" for details). B, Hill plots of the total (E), rapid (å), and slow (Ç) binding.

FIG. 2.
Calcium binding in the absence of AcP at pH 6.23 and 0°C (7). A, total calcium titration curves of the two types of ATPase molecules. Calcium binding to the A form (ϫ), which is in a state of E 1 /E 2 dependent on pH and pre-exists in E 2 at this acidic pH, was obtained by subtracting the binding to the B form (---), which preexists in E 2 independent of pH, from the total binding (E). Calcium binding to the B form was simulated by using parameters of n H ϭ 2.0 and K 0.5 ϭ 5.3 M, obtained on the basis of the observations of the calcium-dependent change in fluorescence intensity of the molecule and maximum binding capacity ϭ half (4.9 nmol/mg of protein) of the capacity (9.8 nmol/mg of protein) of the observed total binding (cf. "Discussion" in Ref. 7). B, Hill plots of the total binding. without the addition of CaCl 2 . Aliquots (1 ml) of the ATPase suspension were placed on the filter. After removal of calcium that was bound to the ATPase by washing it on the filter with EGTA, calcium binding to the calcium-unbound ATPase was initiated by washing the ATPase on the filter with binding solution containing various concentrations of 45 Ca 2ϩ and having the same pH buffer as that used in the preincubation of the ATPase, as described previously (19). The association constants for EGTA-calcium at the employed pH values of 6.23 and 7.40 were 6.31 ϫ 10 4 and 1.38 ϫ 10 7 M Ϫ1 , respectively (20). In the presence of 5 mM AcP, calcium accumulation in the ATPase membrane vesicles was observed. After the calcium that was bound to the ATPase preparation reached equilibrium in the absence of AcP, the preparation further took up calcium at a rate of 2-4 nmol/mg of protein/min when AcP was added. The calcium that was taken up with the help of AcP was completely released by the addition of the calcium ionophore A23187 (data not shown). Therefore, the amount of calcium that was bound to the ATPase in the presence of AcP was obtained by subtracting the amount of the accumulated calcium from the amount of total calcium that was taken up by the membrane vesicles.

RESULTS AND DISCUSSION
We previously showed that two different conformations (A and B) of chemically equivalent Ca 2ϩ -ATPase molecules exist in the SR membrane at a ratio of 1:1 (6). They independently bind two calcium ions (7): the A form binds the ions noncooperatively, whereas the B form binds the ions cooperatively ( Figs. 1 and 2). Recently, however, these two forms have been shown to result from intermolecular interaction of the ATPase molecules (8). Here, we therefore further examined the independence of the two forms in transporting calcium by hydrolytic coupling of a substrate as an energy source. To examine this point, we compared the calcium dependence of the total hydrolysis activity of the two forms with calcium binding to the forms in the absence of the substrate. The comparison was carried out to elucidate whether these forms independently hydrolyze the substrate due to calcium that is bound to each of the forms. For this comparison, it is necessary to use a substrate that reacts with the ATPase molecules to transport calcium across the SR membrane and that does not have any regulatory effect, other than that as a substrate for the molecules, on their respective calcium binding and hydrolysis. AcP has been shown to support calcium transport (11) and to have no regulatory effect on the turnover of AcP hydrolysis (12,13). However, it is not known whether AcP affects calcium binding. In Fig. 3, we therefore first examined calcium binding in the absence and presence of 5 mM AcP at pH 7.40 and 0°C. As reported previously (7), at this pH and temperature, it is thought that the A form is in E 1 before calcium binding and apparently rapidly binds calcium, whereas the B form preexists in E 2 and apparently slowly binds calcium. Two such calcium binding reactions were observed in the absence of AcP at 10.3 M Ca 2ϩ ; ϳ41% (3.5 nmol/mg of protein) of the calcium (8.5 nmol/mg of protein) that was bound at equilibrium was rapidly bound to the A form, which pre-existed in E 1 , within the dead time (ϳ2 s) of our experiments. The remaining 59% (5.0 nmol/mg of protein) of the bound calcium was slowly and monoexponentially bound to the B form, which pre-existed in E 2 , at a rate of t1 ⁄2 ϳ 4 s. At a lower calcium concentration of 0.12 M, only slow (t1 ⁄2 ϳ 4 s) calcium binding to the B form was observed. These calcium binding reactions were not affected by AcP at all, showing that there is no significant effect of this amount of AcP on calcium binding.
As to the intermolecular interaction of the ATPase molecules, monomer ribbon-dimer ribbon transition in the array of molecules on the SR membrane has been found to coincide with E 1 -E 2 transition of the molecules (9). The transition from E 2 to E 1 has been shown to be the slowest step of the transport sequence in the absence of a higher concentration of ATP than that required for phosphorylation of the ATPase molecule (10). On the other hand, the two forms of the ATPase molecules have been shown to be in different enzyme states before calcium binding dependent on pH (7): at pH 7.40, the A form is in E 1 and the B form is in E 2 , whereas at pH 6.23, both of the forms are predominantly in E 2 . The transition rates of the two forms from E 2 to E 1 were previously shown to be the same, independent of temperature (7,21). Here, therefore, the enzyme states of the two forms before calcium binding were set at E 1 and E 2 and at E 2 only, by setting the pH at 7.40 and 6.23, to drive their hydrolysis reaction asynchronously and synchronously, respectively. Moreover, the experiments were carried out at 5 and 25°C to assay the AcPase activity of the two forms under different conditions of stability of the intermolecular interaction of the ATPase molecules.
As mentioned above, because of the split of the two forms into E 1 and E 2 at pH 7.40, these forms are thought to asynchronously turn over their cycles of AcP hydrolysis reaction at this pH. In Figs. 4A and 5A, the total Ca 2ϩ -AcPase activity, which is composed of the AcPase activities of the two forms, was examined by varying the calcium concentration and temperature at the mentioned pH. At 25°C, Hill plots of the total AcPase activity as a function of calcium concentration were biphasic (Fig. 4B). The two lines of the plots, the slopes of which were ϳ0.9 and 1.7, respectively, intersected near the zero point of the ordinate, suggesting the existence of two different types of AcP hydrolysis reaction with the same level of maximum activities. The calcium concentration for the half-maximum activity (K 0.5 ) was ϳ0.2 M. In a previous paper (7), we showed that at pH 7.40 and 0°C, the A form, which pre-exists in E 1 , noncooperatively (Hill value (n H ) ϳ 1) binds calcium with an apparent calcium affinity (calcium concentration for the halfmaximum binding (K 0.5 )) of ϳ2.0 M, whereas the B form, which pre-exists in E 2 , cooperatively (n H ϳ 2) binds calcium with a K 0.5 of ϳ0.2 M (Fig. 1, A and B). Hill plots of the total binding, which is composed of the binding to the two forms at a ratio of 1:1, were biphasic, with slopes of ϳ0.8 and 1.8. The two lines of the Hill plots also intersected near the zero point of the ordinate. The K 0.5 of the total binding was ϳ0.4 M. This profile of the total binding is very close to that of the total Ca 2ϩ -AcPase activity at 25°C, which is shown in Fig. 4. The profile of the total binding at 0°C was not affected by increasing the temperature to 25°C (data not shown). At 25°C, calcium has been found to be rapidly bound to the ATPase molecules at the same millisecond rate irrespective of their enzyme states (E 1 and E 2 ) before calcium binding (21). These results therefore suggest that at a temperature of 25°C, the two forms equally and independently hydrolyze AcP due to calcium that is bound to each of the forms in a different manner. It was previously shown that at 0°C, the A form, which pre-exists in E 1 , apparently rapidly (t Ͻ 2 s) binds calcium, whereas the B form, which pre-exists in E 2 , slowly (t1 ⁄2 Ͼ 2 s) binds calcium (7). Based on the above discussion of independent AcP hydrolysis by the two forms, it is therefore likely that at a lower temperature, the A form, which rapidly binds calcium, more rapidly turns over its hydrolysis cycle than the B form, which slowly binds calcium. This possibility was examined in Fig. 5A, in which the calcium dependence of the total AcPase activity was studied at 5°C. Almost all of the Hill plots of the total activity were monophasic, with a slope of 0.9 -1.3, although a small deviation from the linear line was observed (Fig. 5B). The slope (0.9 -1.3) of the main portion in the Hill plots is near the n H value (ϳ1) of the binding to the A form, which pre-exists in E 1 , whereas the slope (Ͼ2) of the minor portion is rather near the n H value (ϳ2) of the binding to the B form, which pre-exists in E 2 (cf. Fig. 1B). These results support the above discussion of independent AcP hydrolysis by the two forms at pH 7.40, where the forms asynchronously turn over their hydrolysis cycles.
Figs. 6 and 7 show the calcium dependence of the AcPase activity at pH 6.23. At this acidic pH, as mentioned before, it is thought that the A and B forms synchronously turn over their cycle of AcP hydrolysis reaction because they are in the same enzyme states of E 2 before calcium binding. As shown in a previous paper (7), at this acidic pH, the A form noncooperatively (n H ϳ 1) binds calcium with a K 0.5 of ϳ7 M, whereas the B form cooperatively (n H ϳ 2) binds calcium with a K 0.5 of 2-6 M (see Fig. 2A). At the low temperature of 0°C, both of the forms, which pre-exist in E 2 , apparently slowly (t1 ⁄2 Ͼ 2 s) bind calcium because of their slow transition to E 1 . The total binding profile at this pH, which is composed of the binding to the two forms at a ratio of 1:1, was monophasic, with n H ϳ 1.3 and K 0.5 ϳ 5.8 M (Fig. 2B). Such a profile was not affected by increasing the temperature to 25°C (data not shown), although the ATPase molecules have been found to equally and rapidly bind calcium at a millisecond rate (21). At 25°C, Hill plots of the calcium dependence of the total AcPase activity were monophasic, with n H ϳ 1.4 and K 0.5 ϳ 5.0 M (Fig. 6B). The calcium-dependent profile (n H ϳ 1.4 and K 0.5 ϳ 3 M) of the total AcPase activity at 5°C (Fig. 7B) was almost the same as that at 25°C. These profiles were the same as that of the total calcium binding (Fig. 2B), which is composed of the binding to the two forms at a ratio of 1:1. These results suggest that the two forms equally and independently hydrolyze AcP due to the calcium that is bound to each of the forms, even when the forms syn-chronously turn over their hydrolysis cycles.
These results demonstrate the independence of two forms of chemically equivalent Ca 2ϩ -ATPase molecules in hydrolyzing AcP, irrespective of temperature and the synchronism of their hydrolysis reactions. On the other hand, if the two forms of the ATPase molecules were not independently operating in hydrolyzing AcP, either or both of the following results, which are contrasted with the results actually obtained, would be expected from the experiments carried out here. (i) The calciumdependent profile of the total AcPase activity, which is composed of the activities of the two forms, would not correspond to that of the total calcium binding, which is composed of the calcium binding to the forms at a ratio of 1:1, under some condition(s) of pH and temperature except the condition of pH 7.40 and 5°C. (ii) At 7.40, the calcium-dependent profile of the total AcPase activity would correspond to that of the total binding independent of temperature. Taking into account the monomer-dimer transition of the ATPase molecules on the SR membrane accompanying E 1 -E 2 transition of the molecules (9), the conclusion, mentioned above, indicates that the two types of ATPase molecules independently carry out such monomerdimer transition of the same type of molecules, i.e. a structural unit model in which there are two different pairs of molecules, with each pair consisting of the same type of molecules and in which nonequivalence results from the intermolecular interaction of those molecules. The two-pair model is schematically represented in Fig. 8. The concept that at least two types of ATPase molecules are involved in the calcium transport mechanism first emerged from the studies by Froehlich and Taylor (22). Ikemoto et al. (23)(24)(25) reported the existence of two ATPase molecules that behave differently in sequential steps of the calcium transport reaction. Our results support these earlier reports. FIG. 8. Schematic representation of the monomer-dimer transition of the ATPase molecules in each of the two pairs of the molecules accompanying E 1 -E 2 transition. Open and hatched circles denote the two different types of molecules. Based on a recent report (26) that half of the ATPase molecules cannot be phosphorylated, the counterpart of each pair seems to be dormant.